The Role of the Nose in Staphylococcal Infection

Special Report

A recent issue of Infection Control and Hospital Epidemiology presents new studies of an old concept: Chronic nasal carriage of Staphylococcus aureus relates to subsequent postsurgical infection. These sets of papers emanate from two major groups with long-term interest in staphylococcal infections—one group from the Netherlands and the other from the University of Iowa Hospitals.

The Dutch group, led primarily by Kluytmans, reported on three studies.1 Surgery involved coronary artery bypass, valve replacement, and lung resection. The first is a study of the effect of eradicating nasal carriage of S. aureus on subsequent cardiothorasic surgical-site infections. One group of 928 patients served as a historical control, and a second intention-to-treat group consisting of 752 patients was treated with nasal mupirocin ointment bid for five consecutive days, and 116 patients were not treated.

The nasal carriage rates were similar in the historical control group compared to the treated group (15.1% and 16.0%, respectively). Of those 86 patients with positive pre-operative nasal cultures, only 7% had positive cultures after mupirocin therapy. The surgical site infection rate in the historical control was 7.5% and was comparable to the rate in the not-treated group (7.8%). The SSI rate in the treated group was 2.0%, significantly different (P = 0.0023) from the not-treated group. The rates of isolation of S. aureus from infections in the historical controls (37.5%) were similar to the intervention group (39.7%). Thus, the apparent effect of mupirocin was to reduce infections caused by organisms other than staphylococci. Furthermore, during this study there was no development of mupirocin resistance.

In a cost analysis of the same study, Vanderbergh et al found that SSI costs were $16,878 and mupirocin was 62% effective in reducing SSI infection.2 The cost savings were $16,663 per infection prevented. The cost in Holland of a five-day course of mupirocin was $11, probably much less than the cost in our hospital. The increased costs of an SSI were due mostly to prolonged lengths of stay.

Widespread mupirocin usage has a downside demonstrated by two articles in the same issue. The group from SMBD-Jewish General Hospital in Quebec attempted to control methicillin-resistant S. aureus (MRSA) at their 625-bed hospital. During 1990, the rate of infected or colonized patients rose from 0.2 per 1000 inpatients to 8.2 per 1000 inpatients. Control measures included weekly cultures of the nares of all patients for detection of MRSA. Patients positive for MRSA were treated with intranasal mupirocin tid for the remainder of their hospital stay. In a surprising move, in January 1992, the rates again rose from 2.4 to 6.2 per 1000 mupirocin intranasal mupirocin. This approach eventually controlled the epidemic so in July 1993 selective decontamination was again used.

The investigators were smart to save many of the MRSA isolates over the course of their study. A total of 310 isolates were saved from 231 patients and tested for susceptibility to bacitracin, fuidin, and mupirocin. From 1990 to 1993, the rate of mupirocin resistance increased from 2.7% to 65%. Interestingly, there was a stark biphasic distribution of zone sizes between resistant and susceptible isolates, suggesting that the highly resistant strains had a distinct advantage for persistence and spread.

In a large Brazilian hospital where mupirocin was used to control MRSA, a high rate (63%) of mupirocin resistance was noted. These strains are the first described from Brazil. At a second hospital not using widespread mupirocin, the rate of mupirocin resistance was only 6%. Mupirocin-resistant strains were generally cross-resistant to most other antimicrobials except vancomycin. The mechanism of resistance was not elucidated. Earlier reports found mupirocin resistance associated with a plasmid encoding a modified isoleucyl tRNA.

Some patients develop mupirocin-resistant strains slowly. Kluytmans’ group followed 226 patients on hemodialysis and used mupirocin to decolonize S. aureus carriers, among172 of them who were studied. Of the 172, 39% were carriers. After the initial treatment, 94% were negative at three months and 91% at six months, respectively. The rate of bacteremia in the control group (0.04 per patient year) was significantly (P < 0.001) lower than the control group (0.25 per patient year). No mupirocin resistance was seen.

The Iowa group studied 20 hemodialysis patients for the effect of mupirocin on carriage of S. aureus at both nasal and hand sites.6 Fifteen of 20 patients studied had nasal carriage and 12 of these had simultaneous hand carriage. Intranasal mupirocin eliminated carriage at both sites. Molecular typing of these strains showed that hand isolates often matched nasal isolates. Thus, hemodialysis patients who are nasal carriers most likely contaminate their hands with nasal strains.

Mupirocin was used in another report from the Iowa group to eliminate a small health care worker reservoir associated with a prolonged outbreak of MRSA in the Burn Unit at the University of Iowa Hospitals.7 Many other control measures were used to minimize the outbreak, but it appeared that decolonization using mupirocin of a surgical resident and a nurse who cared for eight of the 10 patients brought an end to the outbreak.


Topical antimicrobials have been used for centuries to reduce local bacterial inocula and treat cutaneous infections. Their use has been fraught with the development of resistance (e.g., the introduction of topical gentamicin for therapy of burn wound sepsis resulted in such a rapid emergence of resistance that the product was pulled from the market several years after its introduction.)

The use of topical mupirocin (Bactroban) for impetigo and local staphylococcal infections has not witnessed as rapid an emergence of resistance, although there are few studies searching for mupirocin resistance in non-outbreak settings. In fact mupirocin resistance was most prevalent when it was used for decolonization of all admitted patients to a large Brazilian hospital as reported by Netto dos Santos et al in the midst of large upswings of infection and colonization due to MRSA.4

At the same time, the work from the Netherlands and other work presented in abstract form (Infec Dis Alert 1996;16:45-46) have clearly shown that elimination of staphylococci from the nares of surgical patients reduces the rate of surgical site infections (SSI). If we apply the fiscal analysis of Kluytmans simply to cardiothorasic surgery performed in the United States, there would be a huge potential savings.5 The cost basis used by the Dutch was U.S. dollars in 1991, which certainly has seen about a 15% escalation in the medical marketplace.

So, as for topical nasally applied mupirocin, the issue is how to balance the benefits of decolonization with the risk of losing the efficacy of the product. Hospital epidemiologists will need to define the goals of a decolonization program. The goals may be to reduce SSI in one or more surgical procedures; to eliminate specific carriers of susceptible or resistant staphylococci; to modify hospital-wide problems of MRSA or a raging MSSA clone; or to minimize spread in the community. Each of these goals will have a benefit:risk ratio peculiar to the hospital or geographic region based on the severity of illness in the population at risk, the number of pathogenic staphylococcal clones, or the frequency of mupirocin resistance.

At this juncture, my inclination is to recommendation presurgical decolonization of nasal S. aureus, particularly MRSA strains, in those patients undergoing elective surgery involving prosthetic implants, cranio-facial manipulation or CABG. I would extend these recommendations to reduce or eliminate carriage in surgeons of nasal MRSA. Surgeons need to understand that strains of MRSA can be traced to specific carriers and patients should not be unduly exposed to potential contamination by multiresistant staphylococci carried into the operating theater by surgeons. Dialogue on the local level between surgeons, infectious disease physicians, and hospital epidemiologists can ease the threat that decolonization poses to surgeons and other operating room health care workers. As Boyce suggests in an editorial addressing the articles in this issue of ICHE, populations at continuing risk may benefit from periodic screening with the aim of intermittent decolonization.8

Finally, patients undergoing hemodialysis or peritoneal dialysis who carry S. aureus in their nares may clearly benefit from decolonization. Part of the benefit suggested by the study by Boelaert et al is that hand carriage is reduced by the nasal application of mupirocin.

As more insight is gained about the dynamics and epidemiology of nasal carriage of staphylococci, new strategies will evolve to complement use of topical antimicrobials to reduce or possibly to eliminate permanently the nasal staphylococcal carrier state.


1. Kluytmans JAJW, et al. Reduction of surgical site infections in cardiothoracic surgery y elimination of nasal carriage of Staphylococcus aureus. ICHE 1996;17:780-785

2. VandenBerg, et al. Cost-effectiveness of perioperative mupirocin nasal ointment in cardiothorasic surgery. ICHE 1996;17:786-792.

3. Miller MA, et al. Development of mupirocin resistance among methicillin-resistant Staphylococcus aureus after widespread use of nasal mupirocin ointment. ICHE 1996;17:811-813.

4. Netto dos Santos KR et al. Emergence of high-level mupirocin resistance in methicillin-resistant Staphylococcus aureus isolated from Brazilian university hospitals. ICHE 1996; 17:813-816.

5. Kluytmans JAJW,, et al. Elimination of nasal carriage of Staphylococcus aureus in hemodialysis patients ICHE 1996;17:793-797

6. Boelaert JR, et al. Nasal and cutaneous carriage of Staphylococcus aureus in hemodialysis patients: the effect of nasal mupirocin. ICHE 1996;17:809-810.

7. Meier PA, et al. A prolonged outbreak of methicillin-resistant Staphylococcus aureus in the burn unit of a tertiary medical center. ICHE 1996;17:798-802.

8. Boyce JM. Preventing Staphylococcal infections by eradicating nasal carriage of Staphylococcus aureus: Proceeding with caution. ICHE 1996;17:775-779.